ReMi: A Random Recurrent Neural Network Approach to Music Production

TL;DR

ReMi introduces a novel approach using randomly initialized recurrent neural networks to generate musical elements like arpeggios and oscillations, enhancing creativity with minimal data and computational resources.

Contribution

The paper presents a new method leveraging random RNNs for music production, reducing data needs and computational power compared to traditional AI music generation.

Findings

Generates rich, configurable musical patterns

Requires no training data

Uses less computational power than end-to-end models

Abstract

Generative artificial intelligence raises concerns related to energy consumption, copyright infringement and creative atrophy. We show that randomly initialized recurrent neural networks can produce arpeggios and low-frequency oscillations that are rich and configurable. In contrast to end-to-end music generation that aims to replace musicians, our approach expands their creativity while requiring no data and much less computational power. More information can be found at: https://allendia.com/